Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P616 Early infliximab drug levels predict perianal fistula response to treatment

Y. Davidov1, B. Ungar1, O. Haj Natour1, M. Yavzori1, Y. Chowers2, R. Eliakim3, S. Ben-Horin3, U. Kopylov*1

1Sheba Medical Centre, Gastroenterology, Tel Hashomer, Israel, 2Rambam Health Care Campus, Haifa, Israel; Bruce & Ruth Rappaport School of Medicine, Technion Israel Institute of Technology,, Gastroenterology, Haifa, Israel, 3Chaim Sheba Medical Centre, Gastroenterology, Ramat Gan, Israel

Background

The association of infliximab (IFX) trough levels with luminal clinical and endoscopic outcomes in inflammatory bowel disease is well established. However data regarding the association of perianal fistula response with IFX levels and anti-IFX antibodies are scarce. The aim of the study was to establish whether early IFX TLITL and ATI (at week 6, before 3rd infusion) are associated with perianal fistula response.

Methods

We included consecutive CD patients with perianal fistulas that were treated with IFX between 2008 and 2015. Perianal fistulae were characterised using Parks classification. Response was defined as an improvement in fistula drainage or fistula closure at 14 weeks after first IFX infusion. Infliximab trough levels and ATI were measured using a validated ELISA assay. Patients with unavailable ITL/AI measurements and/or missing clinical follow-up at week 14 were excluded.

Results

In total, 23 patients were included (10 [43.5%] male, mean age of 29.5 ± 23.4 years; 81.9% of the patients were anti-tumour necrosis factor [TNF] naïve) and 91.3% received a concomitant immunomodulator. Perianal fistulae responded to treatment in 8 (34%) of the patients at week 14. The mean ITL at 6 weeks level in responders was 25.6 ± 18.8 vs 6.3 ± 5.9 mcg/mL in non-responders (p = 0.013). The area under the curve (AUC) for ITL at week 6 for prediction of fistular response was 0.83. ITL levels >11.2 mcg/mL at week 6 were associated with fistular response at week 14 with a specificity of 89%. The prevalence of AI was similar between responders and non-responders. ITL and presence of AI at week 14 were not associated with fistular response at week 14. Other clinical and demographic characteristics including concomitant treatment with immunomodulators, Parks classification, and utilisation of setons were similar between the groups.

Conclusion

Early (week 2) infliximab levels predict perianal fistulae response to treatment at week 14. If validated in a larger prospective study, our findings might guide anti-TNF treatment in patients with perianal CD, and support decisions regarding switch or discontinuation of treatment in perianal non-responders to IFX.